MODEL ADOPTION #4171: REJECT — LEDGER BENCH: Apply filter: Session = LONDON_OPEN — improves win rate from 32.6% to 36.2% — E8 Intelligence Research

DECISION: REJECT ELEMENT: No single rule survives — the LONDON_OPEN filter is a session heuristic, not an E8 geometry timing constraint. It contradicts the model's core rhythm logic (Gods-Clock) which already encodes session boundaries via phase transitions. Adding a hard filter would double-count time structure and likely discard high-volatility setups that E8 intentionally targets during Asian/New York overlap. EXPECTED IMPACT: +0.0pp win-rate in live — the +3.6pp backtest edge is likely an artifact of n=356 with regime clustering (summer thin liquidity). The live window's neutral PnL delta confirms no real profit gain. Over 3-6 months, the edge would decay to <1pp due to session-boundary drift and slippage, while increasing missed-trade opportunity cost. RISK: Overfitting to a single session; wkjudas and rizzy5m/7m are not session-designed — they rely on momentum bursts that occur across all hours. Filtering to LONDON_OPEN would reduce trade frequency by ~40%, starve the model

Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com